Two years ago, premium bodywear brand Wolford was on the verge of upgrading its e-commerce site. But with consumer behavior changing rapidly in the midst of the pandemic, its newly arrived Global Director of IT & Digital, Rainer Knapp, called for a rethink. He recalls:
The idea was just to replatform the webshop with the new version of Salesforce, and I stopped that project. Because we said, just doing what we always do — replace the main webshops — will not bring us where we want to be.
Direct-to-consumer has always been a key part of Wolford's business, with a network of brick-and-mortar stores serving consumers eager for its fashionable tights, bodywear and lingerie. Direct online sales surged in the pandemic from around 10% to nearly a third of the total. Knapp decided it was time to start with a clean sheet and plan around what customers would expect from a truly omni-channel experience. He comments:
We said, just a new webshop will not bring us where we want to be. So if it comes to omni-channel, one has to know that you have to think it holistically. It's not just doing click-and-collect — and then most of the companies stop because it gets complicated quite quickly.
Knapp had put that to the test by doing his own mystery shopper experiment when he was about to join the company. He bought six items online, and then walked into one of Wolford's largest stores and asked to return three of them. He recounts:
I entered the store and said, 'Hello, I bought this online, and I would like to return it.' The store associate looked at me and said, 'No, you cannot.'
It's not just the lack of joined-up systems and processes that prevents the physical stores from acting as an extension of the webshop, and vice-versa. There also needs to be reset of processes and incentives. Knapp explains:
They are incented on revenue per square metre. Selling on different channels is not their goal. So it's a complete shift of mindset that you have to do, and that's very important from the cultural point of view. It's not only tech, it's also culture.
Cultural enablement and change therefore has its own dedicated sub-project within the larger project to roll out the new system, introducing store associates to this changed mindset. He elaborates:
Learning the new tools is probably the easiest part. Doing this mind shift and thinking omni in a customer-centric way is probably the more difficult part to learn for them.
Probably you have to change also the KPIs, because if you continue to incent on revenue per square metre, you will never bring them to sell on different channels, regardless. It shouldn't make a difference for them where the revenue is created, but the customer has to be satisfied. That's the main goal.
Making the case for MACH
Taking a greenfield approach to the digital shopping experience led to the adoption of a composable architecture based on MACH principles. The acronym stands for Microservices, API-first, Cloud-native SaaS, Headless and is promoted by the MACH Alliance, an industry advocacy group which held its first global conference in London recently. Knapp was speaking at the event in his role as a volunteer Alliance Ambassador, advocating for education and adoption of MACH.
Based on more than a dozen separate components, the project is now in the middle of being implemented and has been named WolfordX. But before anything could start, it was important to get management buy-in. He recalls:
We had a half a year of evaluation, because we had to make sure also that this new MACH approach was accepted by the business. Which was not an easy job, honestly speaking, because, especially for our supervisory board, coming with tools like CommerceTools, replacing Salesforce, was not easy to explain. In the fashion industry, many people think that Salesforce is the best web shop ever.
Having made the case for the switch, it then went on hold while another project took priority. But even here, the MACH architecture proved useful, as the team was able to use CommerceTools alongside a lightweight business rules engine to provide a capability that had been missed from the original specification. Knapp observes:
It was a use case that was not foreseen. But the flexibility of this architecture and the openness, the API-first approach, of CommerceTools allowed us to find a solution.
The most important benefit of the new architecture is the ability to flex rapidly to respond to change in a way that was never possible with the incumbent system. Knapp says:
With monolithic suites, you always have to live with compromises. Usually, you had to know all your requirements up front, for two or three years, in order to be able to choose the best solution — because once chosen, you are in a vendor lock-in. So reacting then later on becomes quite difficult.
With MACH, it's the opposite. It allows us to introduce new solutions all the time, replace something that's not the best fit maybe, use new microservices. I think we have the possibility, really, with this new design, to accomplish all the requirements our business have as ideas.
Being such a complex project with a very broad and wide scope, we would never have been able to define all the requirements upfront. Now, every week, a new requirement is on the table. And surprisingly ... we not only find the solution — very often we have multiple options to choose from. So it's more the freedom of choice, it's the flexibility, that you build.
Several new capabilities have been built into WolfordX. There is real-time integration from the CommerceTools webshop to the Tealium Customer Data Platform (CDP). This means that visitor behavior can be tracked and later on matched up with the customer profile once the visitor identifies themselves by, for example signing up for a newsletter, enabling personalized recommendations and offers. Being able to access the CDP as the main source of customer data is a big improvement, as Knapp explains:
One of the main problems in marketing has always been that the customer journey is multi-device. Usually four to five devices are used during a customer journey and you lose the customer, because you don't know who he is on the different devices. With the CDP you have the solution because you can aggregate data into one profile and then publish this data to other systems that use customer data, like the POS system in the store. So you can use data coming in from social channels, for example, in the store, to do clienteling, for example.
Another goal was to have a single promotion engine that covers all channels, so that promotion rules can be defined in one place rather than having one set of rules in the POS system and a duplicate set in the webstore. Separating out the promotion engine also allows for more sophisticated rules, as Knapp explains:
In the checkout, the whole basket is sent to the promotion engine, then all the rules are calculated if they apply or not. Or if multiple rules may apply, in which order, stackable, not stackable, etc. It can be really complex. And then the net price is returned to CommerceTools, and the checkout is just closed.
Appetite for innovation
Other components include Bloomreach, Vertex, Adyen and Hubspot. Integration and agile working are key elements of the architecture, and while development can be outsourced, in-house IT architecture skills are essential to co-ordinate it all. Knapp says:
The main difficulty is how to put together all the pieces. It's not about doing a new headless web shop. The product we're building is the architecture, not a new e-commerce web shop. So that's the real product.
Good architects are rare on the market and it's a competitive advantage if you have it in-house.
While the implementation cost has been similar to that of the planned Salesforce upgrade, the team has projected a lower ongoing operating cost. But the biggest impact on the business is a new appetite for innovation. Knapp sums up:
In the past, you very often had the situation when you approached IT as a business manager that you heard, 'No, not possible.' And now we can ... Using this freedom now to implement whatever comes in mind and makes sense for the business is something that will change the behavior I think a lot in the future.
This goes beyond the familiar notion of business-IT alignment, which is often simply a matter of IT following the lead of business. Now, he says:
We think business all the time, the whole day. I consider myself more being a business manager, honestly, with the advantage of having the lever of IT in his hand, than as a pure IT techie